摘要
目的探讨纹理分析与基于不同感兴趣区域(regions of interest,ROI)动态对比增强MRI在确定胶质瘤等级中的作用。材料与方法搜集病理证实的高级别胶质瘤(high-grade glioma,HGG)(WHOⅢ级、Ⅳ级)及低级别胶质瘤(low-grade glioma,LGG)(WHOⅠ级、Ⅱ级)各40例。所有患者均行常规磁共振成像及动态对比磁共振(dynamic contrast-enhanced magnetic resonance imaging,DCE-MRI)检查。ROI由以下两个区域确定:(1)整个肿瘤区域;(2)肿瘤实性部分。使用非参数Wilcoxon秩和检验比较高级别胶质瘤和低级别胶质瘤之间的纹理特征。结果对于整个肿瘤ROI,HGG和LGG之间的不均匀性值的差异有统计学意义(Z=50.37,P=0.01<0.05)。对于实性部分ROI,不均匀性显示HGG和LGG之间的差异无统计学意义(Z=34.65,P=0.06>0.05)。整个肿瘤的不均匀性参数比实性部分ROI具有更高的诊断准确性。结论基于不同ROI的DCE-MRI纹理分析可以为评估胶质瘤等级提供可靠的价值。
Objective:This study was designed to investigate the role of texture analysis and dynamic contrast-enhanced MR based on different regions of interest(ROI)in determining glioma grade.Materials and Methods:Forty cases of high-grade glioma(gradeⅢandⅣWHO)and low-grade glioma(gradeⅠandⅡWHO)confirmed by pathology were collected.All patients underwent routine MRI and dynamic contrast-enhanced magnetic resonance imaging(DCE-MRI).ROI is determined by the following two areas:(1)the whole tumor area;(2)the solid part of the tumor.Texture features between high-grade glioma(HGG)and lowgrade glioma(LGG)were compared by nonparametric Wilcoxon rank sum test.Results:For the whole tumor ROI,the value of inhomogeneity between HGG and LGG was significantly different(Z=50.37,P=0.01<0.05).For the real part of ROI,the inhomogeneity showed that the difference between HGG and LGG was not statistically significant(Z=34.65,P=0.06>0.05).The whole tumor inhomogeneity parameter performed with better diagnostic accuracy.Conclusions:Texture analysis of DCE-MRI based on different ROI can provide reliable value for evaluating glioma grade.
作者
赵鹏飞
牛广明
王少彧
张华鹏
高阳
ZHAO Pengfei;NIU Guangming;WANG Shaoyu;ZHANG Huapeng;GAO Yang(Affiliated Hospital of Inner Mongolia Medical University,Hohhot 010050,China;SIEMENS Healthineers,Shanghai 201318,China)
出处
《磁共振成像》
CAS
2020年第11期975-978,共4页
Chinese Journal of Magnetic Resonance Imaging
基金
内蒙古自治区科技计划项目(编号:2019GG047)。
关键词
胶质瘤
动态对比增强
磁共振成像
纹理分析
感兴趣区域
glioma
dynamic contrast enhancement
magnetic resonance imaging
texture analysis
region of interest
